Here, we discussed the scopes and implementation issues of artificial intelligence and blockchain for precision medicine. Evidence-based medicine is gradually shifting from therapy to prevention and towards individually tailored precision medicine systems. Where, artificial intelligence can be used to automatically detect problems and threats to patient safety, such as patterns of sub-optimal care or outbreaks of hospital-acquired illness. Artificial intelligence can be used to prevent the issues like drug-interaction, over-diagnosis, over-treatment and under-treatment. It can be used more effectively to solve the problems of antibiotic resistant bacteria.
Artificial intelligence uncovers hidden patterns, unknown correlations, and other insights through examining large-scale various data sets. Genome sequencing is the dominant platform for using artificial intelligence based precision medicine.
Moreover, the clinical implementation of precision medicine requires reliable data sharing. Blockchain technology can provide that facility to implement precision medicine in practice. The greatest strength blockchain is a lack of centralized management and ownership. Blockchain uses distributed digital ledger technology. Here patient can participate in managing their own records. No one can tamper the data. Blockchain-based clinical workflow system can assist healthcare organizations in monitoring entire life-cycle of the transactions.
What is Precision Medicine?
Precision medicine is a new scientific way to treat and prevent illnesses tailored to individual on the basis of a person’s genes, family history, medical history, evidence, lifestyle, and environment. Precision medicine allows scientists and clinicians to predict more accurately which therapeutic and preventive approaches to a specific illness can work effectively in subgroups of patients based on their genetic make-up, lifestyle, and environmental factors. Precision medicine is also known as molecular based personalized, predictive, preventive and participatory medicine. Precision medicine is powered by patient data, health records and genetic codes of patients.
Since the beginning of the Human Genome Project, novel technological developments led to the era of omics sciences. The precision medicine can implemented only by incorporating many diverse types of data, from genomes to microbiomes, with patient data collected by health care providers and patients themselves through electronics health devices. The main advantage of precision medicine is that it will bring more new therapeutic strategies, drug discovery and development, and gene-oriented treatment for individual.
What is Artificial Intelligence?
Artificial intelligence is defined as the branch of science and technology that concerned with the study of software and hardware to provide machines the ability to learn insights from data and environment, and adapt in changing situation with high precision, accuracy and speed.
Generally, AI systems include machine learning algorithms for structured data, such as the classical support vector machine and neural network, and the modern deep learning, reinforcement learning as well as natural language processing for unstructured data.
In reinforcement learning the software learn through reward and punishment system. The program keeps track of when a particular action leads to a reward. The machine tries to repeat rewarding sequences of actions and to avoid less-rewarding ones.
Deep Learning is another subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks. Deep learning methods are based on learning feature hierarchies. Automatically learning features at multiple levels of abstraction allow the system to learn complex functions mapping the input to the output directly from data, without depending completely on human-crafted features.
Artificial Intelligence and Precision Medicine
Artificial intelligence in healthcare and precision medicine has enormous potential. Machine learning algorithms and other optimization algorithms for large scale data can be applied to find meaningful patterns, insights and knowledge from the combination of genomic data, phnotypic biomarker data, self-reported observational lifestyle data and environmental data. Machine learning algorithms are used for analysis and interpretation of radiology images like CT scans, X-rays and MRIs.
Currently, evidence-based medicine is unable to solve the issues like hospital-acquired illness, drug-interaction and antibiotic resistant bacteria. Artificial intelligence with precision medicine can be used for solving issues like hospital-acquired illness, drug-interaction and antibiotic resistant bacteria by developing individual genetics and large scale population based modeling.
Convolution neural network based Multifactor Dimensionality Reduction (MDR) feature construction methods are often used for modeling higher-order feature interactions. It can be combined with expert knowledge-guided feature selector for large biomedical data sets.
Genome wide association studies (GWAS) are commonly used for detecting associations between single-nucleotide polymorphisms (SNPs) and common diseases such as heart disease, diabetes, auto-immune diseases, and psychiatric disorders. Artificial Intelligence for Genomics sequencing help medical professionals to interpret how genetic variation affects metabolism, DNA repair, and cell growth. Machine learning algorithms are designed based on patterns identified in large genetic data sets.
AI for the Analysis of Big Data
Healthcare is one of the most data-rich industries, driven by digital health, radiology images, and widespread electronic health records (EHRs) adoption. Data from entire patient populations can be analyzed using AI to discover new evidence and determine best healthcare practices.
AI for Lifestyle Impact Analysis
By observing and analyzing our daily behaviors such as diet, sleep, working style, time management, social interactions and individual’s preferences, we may monitor our daily physical, chemical and mental stress. Machine learning algorithms can learn user’s behavioral trend and analyzing the stress state.
AI for Genome Sequencing
Machine learning techniques are most useful for linking genotype with phnotype, behavior and environment. Machine leaning technologies can be applied for public personal genome data set to learn the genetic risk for different individuals and groups. AI learning systems can be develop for sophisticated prediction algorithms for incorporating multiple genes and multiple environment factors.
What is Blockchain?
Blockchain is a peer-to-peer distributed ledger technology and has three major components:
1. Distributed network: The decentralized P2P architecture has nodes consisting of network participants, where each member stores an identical copy of the blockchain and is authorized to validate and certify digital transactions for the network.
2. Shared ledger: The members in the network record the ongoing digital transactions into a shared ledger. They run algorithms and verify the proposed transaction, and once a majority of members validate the transaction, it is added to the shared ledger.
3. Digital transaction: Any information or digital asset that could be stored in a blockchain could qualify as a digital transaction. Each transaction is structured into a ‘block,’ and each block contains a cryptographic hash to add the transactions in a linear, chronological order.
Advantage of Blockchain in Precision Medicine
In blockchain technology, there is no central authority, there would be fewer errors and frauds. The AI based blockchain uses cryptographic mathematics, which keeps an accurate record of what’s happened in the past. Every time a piece of data is used, a new code is generated, which is based on all previous activity. Which means that if someone later goes back to edit a previous record—say, to hide the fact that they used a piece of data for a particular purpose—it would mess up every subsequent record and be quickly revealed.
Implementation Obstacles and Limitations
Effectively using machine learning methods requires considerable domain expertise, which can be a barrier of entry for bioinformaticians new to computational AI and blockchain based data science methods. The implementation of precision medicine needs data standardization regarding: (1) defining what to collect, (2) deciding how to represent what is collected (by designating data types or terminologies), and (3) determining how to encode the data for transmission.
The clinical implementation of precision medicine requires worldwide and reliable data sharing, as well as regularly updated training programs. Implementation of precision medicine in clinical practice has been relatively slow despite substantial scientific progress in the understanding of precision medicine. One factor that has inhibited the adoption of genetic data to guide medication use is a lack of knowledge of how to translate genetic test results into clinical action based on currently available evidence. However, the power of modern technologies like Artificial intelligence, Blockchain and EHD can make implementation of precision medicine easy.
The future of healthcare holds great promise for applying AI and blockchain to improve many aspects of healthcare process. Potential benefits include genetic based personalizing treatments to maximize effectiveness, enhanced data security, monitoring population health and outcomes, and discovering new evidence, and new drugs.